CHAPTER 11

REVIEW OF LITERATURE

2.1.History of vegetation mapping

A general overview of vegetation of India was obtained through the forest

type map prepared initially by Champion (1936 ) and subsequently revised by

Champion and Seth (1968). They classified the vegetation based on bioclimatic

attributes. This map prepared at a scale of 1:14,000,000 approximately is

undoubtedly of immense significance for perceiving the climatic relationships,

properties and spatial distribution of different vegetation types in the country.

Preservation of the same thematic information at a larger scale of 1: 1,000,000

was then brought out in Forest Atlas of India (1973).

Although Champion and Seth's ( 1968) system of classification, because of

its simplicity, is widely used, several drawbacks in the classification scheme

adopted and the way mapping was done, have been pointed out.(Puri et al., 1983).

Further, the practical utility of such maps are limited in scope, because the map

represents geographical distribution of potential forest type which is likely to

develop in a region in the absence of any human interference, rather than

representing the attributes of actual vegetation as existing at present. Concept of

potential vegetation becomes somewhat hypothetical where natural vegetation

was dramatically altered in the past. In addition, it is now being increasingly

realised that management of natural vegetation needs to be looked, not in

isolation, but in combination with other alternative landuses in a broad

perspective of integrated resource management. Vegetation/land cover

classification and mapping along these lines have been attempted for southern

11

pan of the country by French Institute of Pondichery jointly with Govt.Dept.

(Legris and Meher-Homji , 1968). The main map at a scale of 1: 1,000,000

depicts, important features of natural vegetation and introduced/transformed land

cover types like, agriculture and plantations. It is accompanied by six inset maps

(at 1:5,000,000 scale) showing other environmental features, like relief geology,

bioclimate, land use, agricultural region and potentia) vegetation types. Such

maps provide more information than the earlier ones mentioned but the scale is

too small to provide detailed information for planning and management. Meher­

Homji (1978) prepared the forest map of Peninsular India in I: 1,000,000 scale

along this line.

2.2. Remote sensing

2.2.1. Role of aerial photographs in vegetation mapping.

Basic idea about the use of aerial cameras for forestry purposes started

during 18th and first half of the 19th century. The inventions of aerial cameras,

different types of lenses, sensitive bases to record the images, the image

processing techniques and the air-born platform developed during this time,

boosted the trend. The practical application of aerial photographs and

photogrammetry depended on the developement of the gelatin emulsion by

Maddox in 1871, and roll film by Eastman in 1885 and air plane by Wright

brothers in 1903. The first aerial photograph from the plane was taken by Wilber

Wright in t909.

The use of air planes in forest stand mapping started in 1919 at Canada.

Wilson (I 920) was the pioneer in this field. In] 924, the Irrawaddy delta in

Burma was mapped from the air (Blandford, t924). Vertical aerial photography in

the 1920's in Burma, Indonasia, Papua-New Guinea, Zambia etc provided the

l2

begin ing of remote sensing in tropical forest lands (Bourne, 1931). The general

principles of remote sensing have been described by Avery (1967), FAO (1973),

Simmonet (1983), Curran ( 1985) and Sharma ( 1986).

In India, the pioneers of the Indian Remote sensing programme were the

two eminent scientists, Vikram Sarabhai and Homi Bhabha. Prof: P.R. Pisharati,

father of Indian Remote Sensing, organised the first successful mission of early

detection of coconut wilt disease in ]970 by Remote sensing techniques using a

Soviet aircraft US equipment and Indian scientist.

In India, however, the survey work using aerial photographs started in

1924 for delta mapping. In Indian subcontinent this technique was first used in

Sri Lanka in ]947. The regular survey started in Asian countries during the

period of I950-60's. Since 1925, aerial photographs are in use in India for

preparation of topographical maps. Since 1965, for Pre- Investment Survey of

Forest Resources, aerial photographs were used for forest type mapping and

inventories in India. Opening of the Indian Photointerpretation Institute (IPT) in

Dehra Dun was a major break through in this field. With the opening of IPT, the

photointerpretation technique was more popularised and subsequently many

organization.like Forest Survey of India (FSl), started photointcrpretation works.

Nowadays airborne recording techniques are not only restricted to aerial

photography, where reflected radiations in the visible and near infra red parts of

the spectrum is utilised, but also other sensors developed subsequently such as

SLAR (Side Looking Air Borne Radar), Thermal infra red line scanner and Multi

Spectral Scanner. These instruments have opened new vistas in application

forestry. These sensor developments, in composition with orbital satellite

technology have given way to new developments in resource evaluation.

13

The application of remote sensing and aerial photography particularly to

India were discussed in detail by Singh (1970), Maselekar (1974) and Tomer

(1976). Tiwari (1977) pointed out the us~oJtratified photo maps to estimate timber.

A comparative evaluation of land use and forest type classification and

mapping, using aerial photographs with conventional ground stock mapping was

carried out by Tiwari (1978). Use of aerial photography in stock mapping in

India was started by Tiwari (1978) and Shedha (1978).

Colour infrared photography became universally popular, although its

potential use in forestry was referred to earlier (Spurr, 1948). Application and

studies of infra red colour photography were also undertaken by many researchers

(Fritz, 1967 and Benson and Sierns, 1970).

Vegetation mapping using 1:10,000 scale B&W panchromatic aerial

photographs has been carried out in the western part of Kanha National Park (Roy

et aI., 1986). Changes taking place in the vegetation cover of Kanha National

Park has been identified on the aerial photographs and mapped by Shetha et al.,

(1986).

Versteegh (1968) prepared forest cover type map of Bastar in Madhya

Pradesh in I:25,000 scale using B&W aerial photographs, Gupta and Abichandini

( 1968) conducted air photo analysis of plant communities in relation to edaphic

zones in the arid zone of western Rajastan. Tomar (1968) prepared a manual for

photointerpretation for tropical forests of Kerala and Tamil Nadu. Aerial

photographs have been used for forest and landuse identification in South, Central

and North India (Tomar, 1969).

14

Tomar (1971) suggested 1:25000 to 1:30,000 scale aerial photographs for

forestry studies in Himalayan region with special reference to Jammu and

Kashmir. Van and Joshi ( 1972) conducted a reconnaissance survey of the forests

in Doon Valley with the aid of aerial photographs and identified about 15 land

cover classes in 1: 60,000 scale aerial photographs. Seth (1972), Van and

Joshi (1972) discussed the scope of photointerpretation in Indian Forestry. Seth

and Tomar (1973) used 1: 60,000 scale B&W aerial photographs for volume class

stratification for forest inventories. Tomar and Maslekar (1974) proposed a land

use/forest type classification for aerial photo interpretation. Tomar and Maslekar

(1974) used aerial photographs for mapping and survey.

FAO (1975) proposed a forest cover monitoring programme for tropical

forests. Maslekar (1977) assessed young teak plantations of Attappadi range,

using 1:10,000 B&W aerial photographs. Shedha (1978) emphasized the role of

aerial photographs in the preparation of forest maps. Tiwari (1978) made a

comparative evaluation of cost, time and accuracy of forest and land use maps

prepared from I: 15,000 B&W aerial photographs and ground stock maps in Tehri

Garwal, Himalayas and found aerial photomap was cost effective and more

detailed information content. Shedha (1981) compared forest maps prepared by

ground and aerial photomethods for Gudem Reserve Forest, Andra Predesh.

Mathur et al.. (1984) correlated soil and land use types of a part of Tarai and

Bhabas forest of Utter Predesh by using aerial photographs.

Madhavanunni et al., (1985) prepared forest cover map of a part of

Godavari basin using B&W aerial photographs of 1:25,000 scale. Rekha Ghosh

(1989) prepared drainage map of Eastern India from aerial photographs. Menon

(1988) used 1:15,000 B&W aerial photographs for vegetation mapping in

Attappadi region, Kerala. In this, more than 20 units were mapped. Porwal and

15

Roy <.1991) used I: 10,000 B&W aerial photographs for mapping Kanha National

Park, Madhya Pradesh. Five sublandscapes and physiographic units, four forest

vegetation types, four crown closure classes, under storey components etc were

also distinguished in the study.

Fang (1980) discussed the use of aerial photographs and Landsat

imageries in forest inventory in China. Ibrahim et al., (1986) discussed the use of

aerial photographs in Malaysian forestry. Ibrahim and Hashim (1990) mapped

and classified mangrove forests by using 1:40,000 aerial photographs in Malaya.

Tiner (1990) used high altitude photographs for inventorying wet lands in the

United States. Aerial photographs of two different dates were used to monitor

changes in Sinharaja forest. Srilanka (Banyard and Fernando, 1992).

2.2.2. Role of Satellite imagery in vegetation mapping

The applications of satellite imagery tor forestry related purpose is

relatively new. The era of satellite remote sensing began with the launching of

Larfsat I in July 1972 by NASA, United States. The first Indian Remote Sensing

satellite (lRS lA) was launched in April 17, 1986. Recent research indicates that

optical mechanical scanning is applicable to the identification of forest tree

species ( Olson, 1970).

Quantitative changes in forest cover and effects of fire were successfully

done using satellite remote sensing techniques with special reference to Bandipur

National Park and Mudumalai Wildlife Sanctuary, (Madhavanunni et al., 1986).

Prince (1985) has already used satellite data successfully for studying and

monitoring range conditions in Botswana.

16

In 1983, the forest cover maps of India were prepared by NRSA for the

period 1972-75 and 1980-82. Lal et al., (1990) assessed the extent and location of

deforestation in Kodagu district, Karnataka using t :250,000 Landsat MSS data.

The application of IRS IA data in forestry was discussed by Madhavanunni

el al.. (1991). Porwal and Roy (1992) used 1:50,000 Landsat TM Fee for

delineation and mapping of heterogen. DUS forests of Western Ghats, Kerala and

estimated as overall accuracy as 88.33 percent.

Roy et al., (1992) used 1:50,000 Landsat TM Fee for mapping Chaneka

Wildlife Sanctuary, Ores and compared it with aerial photomap. Application of

remote sensing for ratten resources survey was done by Nandakumar and Menon

(1992). Roy et al., (1993) mapped tropical forests of Andaman islands using

Lands at TM FCe of 1:50,000 scale and identified nine land cover classes.

Habitat assessment of Kaziranga National Park using remote sensing was

studied by Parihar, Panigrahi and Lahar in 1986. Spectral relationship of

grasslands to its biomass in Kanha National Park (MP) has been evaluated (Roy et

aI., 1991).

New remote sensing tool called imaging spectrometer (Airborne

visible/infra red imaging spectrometer AVIRIS) and its application in ecology,

geology and oceanography were described by Greggvane and Goetz, (1993).

Varghese et. al., (1996) prepared bamboo stock mapping using remote

sensed data. Suraj et. al., (1996) prepared a land cover map of a part of

Chimmony Wildlife Sanctuary, Kerala. Mapping of high altitude shola grasslands

of some part of Eravikularn National Park were carried out by Sureshbabu et al.,

(1997).

17

2.2.3.Digital image processing

For digital mapping Landsat data were used by Dodge and Bryant in 1976.

The area of hard wood/soft wood and total forested area were compared with

existing records. Bryant et al., (1980) used Landsat digital data for forest

mapping and compared it with aerial photographs. Computer classification of

data from Landsat has resulted in measurements and maps of forest types tor two

New Hampshire Countries (Arthur and Emily, 1976). Studies on spectral

separability of cover classes were done by many researchers. Spectral

separability analysis of various tropical forest cover classes as recorded on

Landsat MSS data were carried out for two test areas of Northeastern India

(Ashbind Singh, 1987).

Adeneyi (1985) prepared land cover map of a semi arid area of Nigeria

using Landsat digital data. Skidmore et al., (1987) used digital Landsat data for

forest mapping in Australia. In India, digital mapping were carried out by several

workers. Kachhwaha (1983) used Landsat digital data for forest mapping. Singh

and Khan (1989) used digital data for change detection studies. Ashbindu Singh

(J990) integrated digital data with ancillary data to improve supervised

classification. Menon and Sashidhar (1990) evaluated different digital techniques

for land cover mapping. Menon (1991) mapped rubber area using IRS data.

Menon and Ranganadh (1992) used IRS data for mapping Silent Valley region.

Vegetation indices from AVIRS data were used to evaluate spatial

patterns of vegetation type, productivity and potential physiological activity by

John et aI., (1993)

1&

2.3. Vegetation studies.

Ecology began to be formalised as a discipline in Europe in the late 1800.

In 1866 E.H. Haeckel coined the term Ecology (Egerton,1962). In 1905 F.E.

Clements published the first American book on Ecology, "Reseaecb metbods in

Ecology". The first formal textbook of Plant Ecology by Weaver and Clements'

was brought out in 1929.

Based on various ecological aspects, ecosystem studies are carried out on

different levels viz: community structure and composition (Synecology), functions

of component species (Autecology), cycle of material components like Carbon,

Nitrogen, Water, flow of energy etc. Primary step in any synecological study is

the classification of vegetation.

Classical reviews of vegetation classification are those given by Stebbing

(1922), Champion (1936), Razi (1955), Chandrasekharan (1962) and Champion

and Seth (1968). Of these, most widely accepted forest classification system is

that of Champion (1936) and Champion and Seth (1968).

Meher-Homji (1984) recognised 14 phytogeographic zones ofIndia based

on bio-climatic parameters, of which 11 are in Peninsular India and three pertain

to the Himalayas and the Andaman and Nicobar islands. The eleven zones

encompass twenty nine vegetation types.

Succession is considered as a process of ecosystem development (Odum,

1969). Tansley (1920) termed succession for the community development.

Clements (1916) advocated the successional theory, which reveals that the

vegetation undergoes a constant orderly change of species in a particular area,

over a period oftime, i.e. vegetation emerges, develops and attairsclimax state.

19

Quantitative sampling methods in small areas or quadrats were introduced

in a few of the earlier studies of American vegetation (Pound and Clernents,

1898). Quadrat sizes for different vegetation types vary accordingly. The minimal

area of quadrat was standardised by species/area curve method (Cain, 1938). The

increase in number of species with increase in area was first scientifically treated

by Jaccard ( 1912) followed by Brawn-Blanquet ( 1932), Cain (1938), Misra and

Puri (1954), Oosting (1956), Misra (1968), Singh et al., (1984), Basha (1987),

Subhash et al., (1987), Pascal (1988) and Ray (1993).

The degree of presence of a species in a unit area (quadrat) is termed as

constancy (Du Rietz, 1930). Raunkiaer (1934) developed the concept of

frequency and formulated the law of frequency based on percentage frequency.

Raunkiaer (1934) developed the frequency analysis, based on the presence or

absence of species in a number of quadrats. Basal area is used as an index of

dominance by Curtis (1959). Density, frequency, abundance, IVI etc. were

worked out by Phillip (1959) and Muller-Dombois and Ellenberg (1974).

The concept of maturity index was developed by Pichi-Sermoli (1948).

Only very few studies have been conducted in India. Maturity index values of

undisturbed and disturbed evergreen, semi evergreen and moist deciduous forests

in Attappadi were worked out (Menon and Balasubrarnanyan, 1985)). Ray et al.,

(1993) found that the tropical forests of Andaman was moderately mature.

Distribution of species is one of the most important aspects of vegetation

that has attracted many ecologists (Fracker and Brischle, 1944~ Whitford, 1948;

Ashby, 1948 and Cole, 1949). Whitford (1948) used the ratio of abundance and

percentage frequency as a measure of contagiousness among the plants. In most

20

of the cases, abundance/frequency value less than 0.025 indicates regular

distribution, 0.025 to 0.05 random and more than 0.05 contagious (Curtis and

Cottom, 1956).

Gleason (1936 & 1939) proposed the continuum concept. Curtis and

Mclntosh (1951) modified this concept. Continuum index value helps in

evaluating the environmental influence over the vegetation. Since the publication

of the Jaccard's formula, it has undergone several quantitative and qualitative

modifications which are widely accepted (Gleasan, 1920: Ellenberg, 1956~

Pandeya, 1961 and Sorenson, 1948).

The concept of community co-efficient was put forward by Jaccard (1912)

in order to compare two plant communities. Jaccard's formula was modified

later on by Gleason, (1920), Sorenson, (1948) and Ellenberg, (1956).

Diversity index derived from information theory, has been widely used in

recent work of ecologists. Shannon Weaner equation (1949) has been widely

applied to quantify anything that came to hand particularly species diversity.

(Simpson, 1949) proposed another index to study the tloristic diversity and

concentration of dominance.

Diversity and dominance In man made forests were worked out by

Srivastava ( 1986). Pande et al., (1988) had conducted a comparative vegetation

analysis of four plantations. They stated that with stand maturity, the higher

density value the higher is the species richness.

Diversity index is higher for tropical forests for young (5.06) and old (5.4)

_. while for the temperate forests, it is between 1.16 to 3.4 (Knight 1975),

Braun, 1950, Monk, 1967 and Risser and Rice 1971).

21

Concentration of dominance values for temperate vegetation ranges

between 0.16 to 0.99 (Whittaker and Nicring, 1965, Rissem and Rice, 1971). In

the case of tropical forests, concentration of dominance is 0.06 (Knight, 1975).

Higher rate of concentration of dominance is due to lower rate of

evolution and diversification of communities in plantation ecosystems. The same

have been reported for temperate vegetation (Connel and arias, 1964 and

Simpson,1964).

Concentration of dominance varied from O. t5 to 0.22 for trees, 0.18 to 0.3

tor saplings and 0.14 to 0.21 for shrubs (Joshi and Tiwari, 1990)~ 0.89 to 0.308

(Joshi and Behera, 199t)and O. 14 to 0.155 (Pathak et al.. 1993) for mixed tropical

forests of Orissa,

Phytosociological works in Kerala are rather scanty. Some of the

important publications are Singh et aI., (1984) for Silent Valley. Basha (1987) for

the evergreen forests of Silent Valley and Attappadi; Pascal (1988) for evergreen

forests of Western Ghats. KFRI (1980) conducted a phytosocioJogical study in

Attappadi reserved forests. Menon and Balasubramanyan (1985) for Trichur

Forest Division and Pascal and Pelissier (l996) for tropical evergreen forests of

southwest India.

SinghaJ et al. (1986) analysed vegetation of woody species of some forests

of Chakrata Himalayas and Singhal and Soni (1989) analysed woody species of

Musoori. Srivastava (1986) studied diversity and eo-dominance of man made

forests . Laxmi et al., (1987) studied Phytosociology on Machhlad subwatershed

ofPauri Garhwal, UP. Singhal and Shanna (1989) worked on Phytosociology of

22

high level alluvial sal forests and Gangetic tropical moist deciduous forests of

Doon Valley.

General nature of vegetation such as density, frequency, abundance and

constancy have been analysed for major forest divisions viz, Gir, Girnar,

Bhavnager and Rajkot Jamnagar divisions of Saurashtra by Menon and Shah

( 1981). Influence of forests on various environmental aspects such as, rainfall,

soil erosion, watershed management, air pollution, wind, climate, energy and

recreational aspects were reviewed by Menon (1982)

In addition to these, many studies have been made on phytosociological

aspects of forest vegetation (Joshi and Behera, 1990; Sushi) Kumar et al., 1991;

Room Singh et aI., t 991; Lata and Bisht, 1991 and Dani et al., 1991)

2. 3.1 High altitude Shola and Grass land ecosystem

Chandrasekharan (1962) classified high altitude vegetation of Western

Ghats as wet temperate forests and montane grasslands. According to Champion

and Seth (1968), the vegetation of montane region of south India can be classified

into Southern montane wet temperate forest (2A/C), Southern subtropical hill

forest (8A/C), Southern montane wet scrub (2Nds I) and Southern montane wet

grassland (2AJDS2). Gadgil and Meher-Homji (1984) classified these forests

under the biogeographic region of wet evergreen forests of Western Ghats and

vegetation type as montane shola.

Even though there is much controversy over the status of shola-grassland

vegetation since 1938, only a very few studies have been conducted on high

altitude shola grass land vegetation with special reference to South India. The

successional status of grasslands have been subjected to much controversy over

23

the years. Ranganadhan (1938) held the view that the grass land of the Nrilgin

representsclima vegetation just like that of shola vegetation, contrary to the

view of subclimax as held by many workers (Champion, 1936; Bor, 1938:

Shankaranarayan, 1958: Gupta, 1960: Gupta and Sankaranarayan, 1962;

Chandrasekharan. 1962; Noble, 1967 and Lakshmanan, 1968).

Shetty and Vivekanandan (1971) made an attempt to document the

vascular tlora of Anamudi and adjacent areas. Rice (1984) studied the ecology

and behaviour of Nilgiri tahr in Eravikulam National Park. Sreekumar and Nair

(1986) attempted to document the grasses of Eravikulam National Park. Easa

(1996) studied prey-predater relationship in Eravikulam National Park.

Srivastava (1994) has shown the re-establishment of the shola forests in

grasslands of upper Palani hills

No phytosociological study has been conducted in Eravikulam National

Park other than Shibu Jose et al.. (1994). According to him grass lands are steady

state vegetation maintained by edaphic features.

24